CAISO is proposing new methods to address “poor-quality data” from some variable energy resources in its markets to improve grid forecasting, the ISO said in a March 9 straw proposal.
California and the West each year rely increasingly on variable resources such as solar and wind, with 93% of California’s energy coming from these two resources on March 16, for example.
Grid forecasting requires accurate data from variable resources, but CAISO has observed significant differences between a resource’s high sustainable limit and its generation data. This discrepancy indicates that some resources are submitting poor-quality data, specifically high sustainable limit data, the ISO said in the proposal.
A resource’s “high sustainable limit” is the generation capacity of the resource after accounting for weather. For example, a 10-MW solar facility that generates 8 MW on a sunny day has a high sustainable limit of 8 MW. But on a cloudy day, the facility might generate 5 MW, so its new high sustainable limit is 5 MW.
High sustainable limit data is associated with co-located variable energy resources and renewable components of hybrid resources. Firm resources, such as gas and nuclear, are minimally affected by weather.
High sustainable limit data is useful for grid forecasting because it is unimpacted by economic and operational conditions, CAISO staff said in the proposal. In CAISO and the WEIM, all variable energy resources providing ancillary services and all hybrid variable energy components must submit high sustainable limit data, staff said.
CAISO uses a resource’s high sustainable limit data to calculate real-time dispatch and pre-dispatch forecasts. If this data is not accurate, the ISO’s real-time forecast accuracy decreases, leading to inefficient dispatch by overestimating or underestimating a resource’s true generation potential, staff said.
Because data quality is inconsistent across resources, the ISO cannot systematically correct for poor data quality, they said in the proposal.
In CAISO’s initiative on the subject, some stakeholders suggested the ISO — rather than the resource owner — calculate a resource’s high sustainable limit. But CAISO noted that shifting calculation responsibility would not address the overall concerns with data quality.
“An ISO-calculated high sustainable limit would still rely on the quality of data provided by the resource,” staff said. “In addition, the ISO does not have visibility to all operational factors that could impact high sustainable limit, such as the status of an inverter.”
To improve high sustainable limit data, CAISO proposed to provide a new, clearer definition of what good quality data looks like. The ISO provided also an example methodology for solar resource owners to calculate their high sustainable limit data more accurate, though the example methodology serves as “guidance,” but other ways of calculating high sustainable limit data could meet data quality requirements too, staff said.
Stakeholder feedback on CAISO’s straw proposal is due March 30.
The announcements include expansion of the U.S. LNG sector, procurement of small modular reactors (SMRs) and a new U.S. coal-fired power plant.
The Indo-Pacific Energy Security Ministerial and Business Forum brought representatives from 17 countries to Tokyo March 14 and 15. U.S. officials framed the results as a boost for the American workforce and a step toward President Donald Trump’s vision of U.S. energy dominance.
The list of agreements announced by the U.S. Department of the Interior (DOI) is lengthy due to the number of deals included rather than the level of detail provided.
Some of the parties involved elaborated in their own announcements. Others did not.
Details were minimal on the $1 billion agreement in principle between Terra Energy Center and Hyundai Heavy Industries Power Systems to provide boilers for a new 1.25-GW coal-fired power plant in Alaska. KOREIT committed a $500 million equity investment.
DOI said it was the first utility-scale announcement of its kind since approximately 2006. If it comes to pass, it will prove wrong many experts, observers and pundits who predicted no new coal plants would be built in the U.S. because of regulatory risks under future Democratic administrations and uncompetitive operating costs.
The DOI announcement framed it as part of America’s “Big Beautiful Coal” resurgence, one of Trump’s regular talking points. If a resurgence happens, it would mark the end of a steep and sustained decline: U.S. coal power generation dropped from 2,016 TWh in 2007 to 652 TWh in 2024 as plants were run less often or retired in the face of stricter emissions controls, cheaper natural gas generation and proliferating renewables.
The most recent new U.S. coal plant was the 900-MW Sandy Creek Energy Station in Texas, which began construction in 2008 and started commercial operation in 2012.
Other announcements ranged from letters of interest to binding commitments covering a wide range of technologies. LNG figured prominently in the DOI announcement and in the $56 billion tally:
Venture Global closed an $8.6 billion final investment decision for the CP2 LNG facility in Louisiana, bringing total financing to $20.7 billion.
Venture Global and Hanwha Aerospace executed a 20-year purchase agreement valued at about $10 billion for 1.5 million tons of U.S. LNG per year starting in 2030.
X-Energy and Doosan Enerbility struck a binding agreement to manufacture 16 main power systems for X-Energy’s Xe-100 SMR and to build the world’s first dedicated fabrication facility for the reactors.
Holtec International, Mitsubishi Electric and Hyundai Engineering & Construction entered a memorandum of understanding to jointly deliver the first two units of Holtec’s SMR-300 in Michigan and to deploy more SMR-300s across the Indo-Pacific region.
GE Vernova and Hitachi agreed to advance market development and commercial opportunities for deployment of their BWRX-300 SMR in Southeast Asia.
Other announcements included:
LG Energy Solution and Tesla reached a supply agreement to build a $4.3 billion lithium-iron-phosphate prismatic battery cell factory in Michigan that will supply Tesla’s Megapack 3 energy storage systems.
The U.S. and South Korea are exploring a critical minerals memorandum of understanding.
The U.S. Trade and Development Agency (USTDA) awarded an unspecified grant to PT Geo Dipa Energi supporting a pilot project to assess viability of U.S. ion-exchange technology from Lilac Solutions.
The USTDA organized the summit. The U.S. delegation was led by Interior Secretary Doug Burgum, who also is chair of the National Energy Dominance Council (NEDC); Environmental Protection Agency Administrator Lee Zeldin; EXIM President John Jovanovic; NEDC Director Jarrod Agen; and USTDA Deputy Director Thomas Hardy.
The digital world may be driving much of the growth in electricity demand, but physical limits are shaping how the industry responds. And few limits are more apparent than the shortage of transformers.
Expanding manufacturing and extending the life of existing transmission assets should be considered as essential to the country’s economic future as the build-out of the grid is. And like grid expansion, easing the shortage needs to be a national priority. Yet it’s hard to create priorities without a clear picture of the future grid.
At a recent Silicon Valley media summit, Anthony Allard, executive vice president and North America region head at Hitachi Energy, called for national planning, “like we see in some countries around the world: 20, 30 years.” That forward view is essential for infrastructure planning. “We need to take that long-term view to make sure that we optimize the grid and infrastructure that we build in the country.”
Across North America, utilities report load growth numbers that would have seemed implausible a decade ago. Gigawatt-scale artificial intelligence data centers are opening anywhere that access to generation or the grid is relatively easy. Manufacturing is reshoring. Transportation and building heat are electrifying. Some regions are seeing year-over-year load growth in the mid-single digits, a dramatic shift after decades of flat demand.
All of that new generation and load has to pass through a narrow physical choke point: the transformer fleet.
Transformer availability is a bottleneck choking the growing market, with lead times for power transformers running at 128 weeks and for generation step-up transformers at 143 weeks, according to a Q2 2025 Wood Mackenzie study. While interconnection queues may be longer, this transformer shortage prevents the market from working around them: without transformers, hyperscalers won’t be able to end-run the queues with behind-the-meter generation.
Growing Demand, Meet the Aging Grid
The growth in demand is far from the only issue. It is happening against the backdrop of supply chain challenges and fluctuating tariffs that have muddied the math for companies considering manufacturing facilities in the U.S. And it’s happening at a time when the aging grid requires more upkeep than ever.
Dej Knuckey |
The grid is old, and there are plenty of operational transformers living on borrowed time. A Bank of America Institute study found that 31% of transmission infrastructure was within five years of, or beyond, its useful life. “Further, 67% of electric utilities’ spending in 2024 was on replacements ($63 billion), while only $32 billion was spent on new lines and substations.”
For the small transformers that hang on the poles near your home, that’s one thing: They are uniform and fairly easy to replace, even if supplies are short. It’s the behemoths, the ultra-high voltage power transformers, that are the issue: Without those, the grid simply can’t grow.
Upgrading the grid is an urgent national issue, so much so that federal grid upgrade grant programs have survived this administration while almost none of their renewable counterparts have. But it’s the companies that control the transformer manufacturing base that are setting the pace of the energy transition.
The Workhorse Becomes the Constraint
Transformers change voltage, stepping it up for long-distance transmission and down so it can be delivered safely to cities, factories and data centers. Without them, electrons don’t move from generator to load.
For most of the past 20 years, transformers were an afterthought in planning conversations. They were heavy, custom pieces of hardware with long but manageable procurement timelines. They rarely drove interconnection queues or delayed major industrial projects.
That is no longer the case.
Large power transformers, the 800-kV UHVDC ones essential for efficient long-distance power transmission, are custom-designed and built, often the size of a two-story home and weighing as much as 400 tons. They are engineered to specific customer requirements, and, surprisingly, the experts I met with were aware of only one customer standardizing across its fleets of transformers.
To date, there is no industry-wide effort to share specifications that could help bring transformers to market faster. Perhaps they should look to other industries, like the airlines, which build similarly large and complex machines: a certain amount of standardization has made the industry more efficient while still allowing for customization. There’s no need to go as far as Southwest Airlines’ famous “only Boeing 737” policy, but it would be worth exploring whether unifying around specs could help manufacturing scale as quickly as possible.
A Supply Chain That Can’t Simply Scale
The current load surge driven by AI, electrification and reshoring is structural, not cyclical. If that growth persists, North America will need smarter use of existing infrastructure and a sustained expansion of manufacturing capacity.
It’s tempting to assume high demand will naturally attract new entrants. But transformer manufacturing is not easily replicable.
Standing up a new factory takes years and hundreds of millions of dollars. Even expansions of existing facilities require long lead times and a deeply skilled workforce that understands high-voltage design, insulation systems and quality control. It is not software; it is steel, copper, oil, insulation and people.
The policy challenge is that transformer factories require long-term visibility into demand. Capital-intensive facilities are hard to justify in a world of volatile interconnection reforms, shifting trade rules and uncertain permitting timelines. Manufacturers need confidence that today’s demand spike is not tomorrow’s policy whiplash.
Hitachi Energy’s Billion-dollar Bet
Into that gap steps the largest global supplier of transformers and other high-voltage equipment: Hitachi Energy.
In September 2025, the company announced it would expand production of critical electrical grid infrastructure in the U.S., with its largest $457 million investment in a new power transformer facility in South Boston, Va. The plant expands an existing distribution and medium-voltage (up to 345 kV) transformer production facility.
Before its $457 million expansion, the plant in South Boston, Va., produced distribution and medium-voltage transformers. | Hitachi Energy
Construction is expected to be complete in 2028, with deliveries beginning in 2029. It will create more than the 800 new jobs, on top of roughly 700 existing employees at that location. The announcement signals relief is coming — but not immediately.
The announcement included additional investments in a transformer components facility in Tennessee ($106 million) and expansion of its dry-type transformer capacity in Virginia ($22.5 million) and its high-voltage components production in Pennsylvania ($70 million).
Investments like these are not a slam dunk, despite years of capacity already accounted for. While demand is surging now, the production line will take years to come online and will take even longer to pay back the considerable capital investment. The “super-cycle” in the market is likely to continue for at least the next five to 10 years, Arya Barirani, CMO of Hitachi Americas and Hitachi Digital, told me at a recent briefing. The biggest question that companies like Hitachi Energy have to contend with is: What happens when that demand is no longer there or not at the same clip?
For RTOs and utilities, that timeline matters. Transmission planning cycles and interconnection reforms are operating on similar multiyear horizons. A transformer factory that delivers its first units in 2029 aligns more with the back half of the decade’s capacity needs than the immediate surge.
Who Gets a Transformer?
In a constrained market, allocation matters. Large, creditworthy customers, particularly hyperscale data center developers, may be better positioned to secure production slots by committing capital early. Smaller or rural utilities could find themselves further back in the queue.
The market’s demand is evident in the fact that Hitachi Energy’s UHVDC transformer production capacity already is reserved for several years, though they would not share specific numbers.
This dynamic has implications for RTOs and state regulators. If transformer availability becomes a gating factor for interconnection, market rules and cost-allocation frameworks increasingly will shape who moves forward and who waits.
The industry already is experimenting with approaches in which large customers pay up front for network upgrades. In a world of scarce transformers, those financial signals could directly influence manufacturing schedules.
Optimizing the Grid We Have
Expanding manufacturing is only half the story. The other half is lifecycle management.
Transformer fleets across North America are aging. Even with perfect condition monitoring and predictive analytics, transformers have finite thermal and mechanical limits. Insulation ages. Bushings fail. Core steel saturates. At some point, the grid needs more iron in the ground.
Many large units are decades old but not yet at end of life. Extending their useful service through refurbishment, monitoring and predictive maintenance can free up scarce capital and defer replacements.
“The fastest way to add capacity is to really take care of the existing infrastructure and the grid,” said Emrah Ercan, Hitachi Energy’s vice president and head of service for North America. So, as other parts of Hitachi Energy are “adding steel” to the grid, Ercan’s team’s mandate is “to take care of the existing grid and get as much as possible as we can out of the transformers [and] the high voltage assets that we have on the ground.”
Large power transformer mounted on a wide trailer ready for transportation from Hitachi Energy’s production facility in Varennes, Quebec | Hitachi Energy
The company is the leader in the space, with $230 billion of existing HV assets deployed globally, $60 billion of which are in the U.S. By using its digital expertise to deliver predictive maintenance, it is well positioned to demonstrate the value of O&M to optimize performance and prolong the life of those assets.
Digital condition monitoring — from dissolved gas analysis to thermal imaging to advanced sensing — allows operators to detect insulation breakdown, overheating or abnormal loading before catastrophic failure. AI-enabled asset management platforms, such as the digital twins discussed in my most recent column, synthesize historical maintenance data, environmental conditions and load patterns to optimize maintenance schedules.
In a world where new units take 30-plus months to arrive, preventing one unexpected failure can be worth millions of dollars in avoided outage costs and reputational damage. But while digital tools and efficiently deployed maintenance teams can stretch capacity, they cannot create it.
What to Watch
For grid planners and market participants, three signals are worth tracking:
New factory announcements and expansions in North America. Capacity geography will influence project timelines and transport logistics.
Regulatory and trade policy shifts affecting electrical steel, copper and high-voltage equipment.
Utility filings and RFPs that explicitly identify transformer lead times as a binding constraint.
The 21st-century grid is trying to move gigawatts of new clean generation and AI-era load through a narrow physical waist: the global transformer fleet. Software can make that waist more flexible. Digital twins and predictive maintenance can squeeze out incremental performance. But only sustained investment in steel, copper and skilled labor will widen it.
Transformers may not be glamorous. But for the foreseeable future, they are the hottest thing on the market.
The companies like Hitachi Energy that are committing billions today, and the regions that attract those factories, will determine how fast the next decade of grid build-out can proceed.
Power Play Columnist Dej Knuckey is a climate and energy writer with decades of industry experience.
Many electric utilities are not prepared for the “unique challenge” presented by the expected rapid growth of data centers and large loads on the grid, and the ERO must support the industry through multiple measures including new reliability standards, according to a report released by NERC.
The new report, published March 17, reviews the responses by registered entities to a Level 2 alert sent in September 2025. That alert provided 105 questions for respondents to answer according to their functions: 44 applied to transmission owners and distribution providers; 23 to transmission planners and planning coordinators; 28 to reliability coordinators, balancing authorities and transmission operators; and 10 to resource planners.
Entities’ responses to the Level 2 alert revealed that utilities expect major changes in the composition of large loads on their systems in the next five years.
In one chart, NERC aggregated entities’ predictions for the demand associated with large loads in 2028 and 2030 as 300.5 GW and 612.7 GW, respectively — significantly more than the 69.5 GW recorded at the end of 2025. About one-third of the demand in 2028, 109.1 GW, is expected to consist of data centers. That share is expected to more than double two years later to 400.8 GW. However, NERC cautioned its confidence in these figures is “questionable due to the differences associated with entity interpretations on the word ‘forecast.’”
NERC also found that “entities largely did not have a formalized definition of ‘large load.’” Those that did revealed a wide range of thresholds, ranging from 100 kW to 400 MW. The greatest number of respondents to this question, which was directed to TOs and DPs, indicated their cutoff was either 50 or 75 MW. About 50 respondents each chose one of these options.
Other popular selections were 10, 20 and 25 MW, each of which was chosen by about 20 utilities. NERC indicated that DPs tended to report thresholds below 20 MW, while all TO respondents reported thresholds above this level.
Recommendations Unfulfilled
The Level 2 alert listed five recommendations related to maintaining reliability with large loads:
TOs should create clear facility design and performance criteria in their interconnection requirements for large loads;
TPs and PCs should establish a comprehensive interconnection and systemwide study process to assess the reliability impacts of large loads;
TOs need a comprehensive load commissioning process that ensures operational readiness for large loads;
TOs should establish operating protocols and communication infrastructure to support reliable ongoing operations after large load facilities enter commercial operations; and
TPs, RPs and PCs should identify and implement a process to include large loads in their near-term and long-term planning horizon demand forecasts.
Responses to the questions showed many utilities have not taken steps to meet these recommendations, NERC wrote, indicating “that additional steps are needed to reliably integrate computational loads.” The ERO is working on several follow-up actions.
First, NERC plans to issue a Level 3 alert on computational large loads by May. Unlike Level 2, a Level 3 alert identifies specific actions deemed essential for certain stakeholders to maintain grid reliability. NERC’s Board of Trustees must approve the issuance of a Level 3 alert.
Second, the ERO expects to draft registry criteria updates to address the integration of computational loads, along with updating reliability standards as needed. These efforts are scheduled to begin in the second quarter of 2026. NERC’s third planned action is to release a reliability guideline in mid-2026 that will “immediately share best practices to improve reliability.”
NERC encouraged registered entities to stay up to date on the initiatives by following its Large Loads Action Plan webpage.
Valley Link Transmission Group Plans 765-kV Line in Virginia
Valley Link Transmission, a partnership between Dominion Energy, FirstEnergy Transmission and Transource Energy, is planning a 115-mile, 765-kV line in Virginia.
The final route of the $1 billion project is yet to be determined, but a map of possibilities shows it could cross nine counties: Appomattox, Buckingham, Campbell, Culpeper, Fluvanna, Goochland, Louisa, Orange and Spotsylvania. PJM has already approved the project.
Officials anticipate the State Corporation Commission will take a year to evaluate the project and its route, with a decision coming around September 2027. The line is expected to be completed in 2029.
The Senate passed a bill that would expand the Public Service Commission from three members to seven, who would represent areas of the state rather than serving in an at-large capacity.
According to the bill, the governor would appoint four commissioners by July 15. Two commissioners would serve two‑year terms while two would serve four‑year terms. The governor would be required to select appointees from a list of three names submitted by the lieutenant governor, house speaker and senate pro tempore for each position. Commissioners elected after June 1 would serve six-year terms.
The bill does not change the way utilities are regulated or address profit but would freeze base rates through June 2029.
Senate Passes Bill Exempting EPIC from Open Records Law
The Senate voted 29-6 to approve a bill that would exempt the taxpayer-funded Energy Planning and Inventory Commission (EPIC) from the state’s Open Records Act.
The bill would exempt “information, records, data, files, documents or correspondence” created by EPIC. The legislature created EPIC in 2024 to review utilities’ plans to retire fossil fuel-fired power plants, along with analyzing supply, demand and infrastructure.
The bill would also remove appointments made by Gov. Andy Beshear to an executive committee within EPIC and give appointment power to Attorney General Russell Coleman.
The town of Harrington voted to enact a moratorium on utility‑scale solar farms.
The Solar Energy Ordinance authorizes personal ground‑mounted systems and personal and commercial roof‑mounted installations but does not authorize power to be distributed to the grid for profit. The commercial provision applies only to businesses that install systems to power their own operations.
House Passes Bill Requiring PUC to Consider Affordability
The House passed a bill that would require the Public Utilities Commission to consider the impact of affordability and publicize data from utilities.
The amended bill directs the PUC to develop an affordability metric to assess the impact of bills on customers’ overall energy burden, which would be submitted to a legislative committee by Jan. 15, 2027. The commission also would be required to publicize data from utilities related to credit and collection activities, and to conduct a review of rates, during which it would consider options to contain costs in delivery rates, reduce transmission and distribution bill volatility, and increase bill transparency.
The Department of Environmental Quality unanimously approved a plan by xAI to build a 41-turbine natural gas power plant.
The generator would power xAI’s data centers in Memphis, Tenn.
The DEQ’s decision comes three weeks after it held a town hall to invite feedback on xAI’s application. Dozens of residents and advocates — both from Mississippi and Tennessee — spoke out against the permit. No one spoke in favor of the facility.
NV Energy Asks PUC to Delay Peak Demand Charge Start Date
NV Energy asked the Public Utilities Commission to push back implementation of its peak demand charge until Oct. 1.
The legality of the demand charge is being challenged by the Attorney General’s Bureau of Consumer Protection. It was scheduled to go into effect April 1. Under state law, utilities can ask to change the implementation of new charges up to 10 days before they are scheduled to go into effect.
South Central Judicial District Judge Jackson Lofgren revoked Summit Carbon Solutions’ permits for underground carbon dioxide storage, saying parts of the state law under which they were issued are unconstitutional.
The decision is the second time a judge has reached the conclusion that the 2009 state law violates the state’s constitution. The law authorizes regulators to permit the storage of carbon dioxide beneath the property of nonconsenting landowners.
A Summit spokesperson said the company is reviewing the decision and evaluating next steps.
Gov. Rhoden Signs Bill Limiting Utilities’ Wildfire Liability
Gov. Larry Rhoden signed a bill that will limit utilities’ liability in wildfire lawsuits.
Once the law takes effect July 1, people who sue will be able to recover damages only from utilities that either fail to file fire mitigation plans and follow them, or act with criminal intent or exhibit “willful and wanton” misconduct. No claims will be valid after four years and plaintiffs will be limited in the types of damages they can recover.
The Public Utilities Commission approved a $3.3 million rate hike for Otter Trail Power.
The company originally asked for $5.7 million, which would have raised average residential rates by more than $14/month. The settlement will raise rates by $8.97/month.
Black Hills Energy also has asked for a rate increase that would raise rates by about $25/month.
The Public Service Commission unanimously approved We Energies’ purchase of the 200-MW Dawn Harvest Solar and Battery Energy Storage Facility.
The $443 million project will provide 150 MW of solar capacity and 50 MW of battery storage. We Energies will own the storage facility and 80% of the solar generation. The remaining power will be split between the Wisconsin Public Service Corporation and Madison Gas & Electric.
The project is expected to be operational in 2028.
BloombergNEF: Global Wind Power Installations Hit All-time High
Global wind capacity additions hit an all-time high in 2025, marking a third consecutive year of record installations, according to BloombergNEF’s “Global Wind Turbine Market Shares 2025” report.
Project developers brought 169 GW of wind online in 2025, 38% more than in 2024. About 161 GW (95%) of the additions were onshore, while 8 GW were installed offshore.
China’s onshore wind sector accounted for most of the growth, becoming the first market to add more than 100 GW in a single year.
The Trump administration is considering waiving the Jones Act for 30 days to ensure energy and agricultural shipments can move freely between U.S. ports, Press Secretary Karoline Leavitt said.
Under the Jones Act, goods shipped between U.S. ports must be carried on U.S.-built, U.S.-flagged and mostly U.S.-owned vessels. The requirement sharply limits the number of tankers available for domestic shipments but is supported by maritime labor unions. The move would be aimed at combating spiking fuel prices and other disruptions since the start of the U.S.-Israeli war on Iran.
Seven maritime labor unions have publicly rejected the idea.
Trump Memo: TVA Employee Comp Should be $500K at Most
President Donald Trump recently issued a memo calling for the TVA board to limit compensation for all employees to $500,000.
The move would affect not just new CEO Don Moul, who is set to be paid $6 million/year, but about 230 other employees. Former board members said TVA would likely struggle to attract a qualified CEO for only $500,000.
The CEO role was created in 2006 after Congress amended the TVA Act. The move also changed the three-person executive board into a nine-person part-time corporate board.
The PJM Market Implementation Committee endorsed an RTO proposal to establish penalties for load management and price-responsive demand (PRD) resources that underperform during pre-emergency deployments.
The penalty would be set at half the rate levied against resources that don’t meet their capacity obligation during a performance assessment interval (PAI), which amounts to about $1,150/MWh based on capacity prices for the 2027/28 delivery year. The formula mirrors the calculation for PAI penalties but doubles the number of expected deployment hours each year to 60. They would count toward the annual stop-loss for Capacity Performance penalties.
PJM’s Pete Langbein said the lower rate reflects that pre-emergency deployments are less severe than PAIs.
Penalty revenues would be awarded to overperforming curtailment service providers (CSPs) if the fleet-wide response meets or exceeds the amount committed. If demand-side resources under-respond, a pro-rated share of the revenues would be allocated to load-serving entities.
The PJM proposal received 86.1% support, while an alternative offered by Voltus received 39.3% and two packages from the Independent Market Monitor received 25.7% and 14.4%.
Voltus
Voltus adopted a similar formula to PJM, but it added a 50% derate to account for the diminished reliability risk associated with pre-emergency events and increased the number of expected deployment hours to 90. The resulting penalty would have been 16.7% of the PAI rate, or about $383/MWh.
Revenues would have been allocated to overperforming demand-side resources at 120% of the penalty rate, pro-rated for the amount they exceeded their assignment. The 20% adder is intended to create an incentive to overperform without allowing a windfall if the bulk of the response is from a small number of CSPs. Any excess would be provided to LSEs.
If the number of pre-emergency load management hours exceeded expectations, the package would have increased the overperformance bonus to 150% to counteract potential fatigue. The possibility of load management deployments becoming more regular as reserve margins tighten has become a frequent subject for stakeholders concerned it could drive away participation.
Monitor
The Monitor’s proposals would have required demand-side resources to curtail according to PJM instructions. Rather than a set penalty rate, it would have withheld daily capacity payments from the latter of the start of the delivery year or their last successful performance or test, spanning to the next successful performance or test. The amount withheld would have been based on resources’ shortfall and the revenues would be entirely allocated to loads rather than to other demand-side resources.
“Load pays for these resources, and load should receive the penalty revenue when the resource fails to perform,” Monitor Joe Bowring said in an email to RTO Insider.
The Monitor’s alternative proposal would have measured performance for each registration and prevented CSPs from netting performance across sites.
David Mabry, representing the PJM Industrial Customer Coalition, argued the Monitor was misconstruing what load management and PRD resources are expected to provide. Rather than provide a set reduction in load, they must maintain their load below a pre-defined level when dispatched.
Bowring said the Monitor’s proposals would prevent resources from being paid for being capable of curtailing when they do not do so when requested.
Pamela Wildstein, a market analyst with the Monitor, said the requirement is to reduce demand when dispatched and cited the tariff provision that states the requirement.
The alternative package was introduced during the meeting to define the performance obligations of demand-side resources with the intention of clarifying what the penalties are for. Bowring said there are several key weaknesses in how PJM defines the obligations of demand resources as capacity resources. Those include “not actually requiring a reduction in load when called to respond by PJM; not measuring the current load and therefore not accurately measuring reductions in load by the resources; simply ignoring actual increases in load by demand resources when called to respond; and allowing aggregation across hours and resources rather than calculating penalties by hour and by individual registration.”
Stakeholders Endorse Quick Fix to Include Batteries in Planning Models
VALLEY FORGE, Pa. — The PJM Planning Committee on March 10 endorsed by acclamation a quick-fix proposal to include battery storage dispatch in the RTO’s planning models.
The revisions to Manual 14B: PJM Region Transmission Planning Process would model storage availability by season to reflect the longer duration of winter events. Batteries are currently modeled as offline in the Regional Transmission Expansion Plan (RTEP) base cases, but they are included in generation deliverability studies.
In presenting the second read of the proposal, Lead Engineer Julia Spatafore said including storage in the RTEP analysis would increase the resources available for transmission reliability, as well as better align regional planning with state policies and the determination of network upgrades for projects in the interconnection queue.
When modeling the system for the summer, the battery availability would be set at the lesser of its effective nameplate capacity (ENC) or the ENC multiplied by the resource’s duration, then dividing that value by 4 and multiplying the quotient by the fleet effective equivalent demand forced outage rate. The 4 in the equation represents the expected duration of summer events; for the winter, it would be replaced with an 8.
Director of Transmission Planning Sami Abdulsalam told RTO Insider that the reliability challenges presented by summer events tend to center around a single peak, while in the winter they tend to span the period between the morning and evening peaks.
He told the committee the change is a “kick start” to the RTO’s effort to use batteries for reliability.
Paul Sotkiewicz, president of E-Cubed Policy Associates, argued the change is too significant to proceed under the quick-fix process, which allows an issue charge, problem statement and proposed solution to be voted on together.
PJM intends to seek endorsement from the Markets and Reliability Committee at its meeting April 22. If approved, the changes would be implemented immediately.
Transmission Expansion Advisory Committee
Supplemental Projects
UGI Utilities presented a $94 million project to serve a 200-MW customer seeking to come online adjacent to the Hunlock Creek substation in Pennsylvania in 2027.
The customer would initially interconnect with 100 MW to be served by upgrades to the 66-kV bus at the Mountain substation. The second half of the load would come online in 2029 following the construction of a 230/66-kV substation named Newport, which would cut into the 230-kV Susquehanna-T10 line and connect to a switching station to serve the customer with two 66-kV lines. The project is in the engineering phase with a projected in-service date of Sept. 30, 2029.
UGI canceled a $33 million project to serve a 384-MW customer near Nanticoke, Pa., because the customer canceled the project. The upgrades would have included constructing a 230-kV substation, which would have cut into two 230-kV lines between Mountain and Susquehanna.
PPL presented four new service requests to serve large loads in Pennsylvania, each exceeding 1 GW. The requests seek to:
bring a 200-MW load to Washingtonville in 2028 and ramp to 1,500 MW by 2032;
site a 100-MW load in Archbald in 2028 and ramp to 1,200 MW by 2033;
construct a 75-MW load in Berwick in 2028 and grow to 1,500 MW by 2033; and
develop a 60-MW load in Nescopeck in 2028 and ramp to 1,350 MW by 2032.
PPL’s Robin Lafayette said direct connection costs, such as new substation equipment and lines to serve the new load, are currently fully allocated to the customer, while the costs of upgrades to existing facilities are assigned to the transmission zone in which the load is located.
Dominion Energy presented a $115 million project to address a 300-MW load drop violation identified in a 2029 do-no-harm analysis. The project would tear down and rebuild the 19-mile, 230-kV lines between the Gordonsville, Louisa CT, South Anna, Desper and Foxbrook Lane substations in Virginia. A parallel line would be constructed connecting Gordonsville directly to the Wesbey Drive substation, bypassing the other substations. The project is in the planning phase with a projected in-service date of March 1, 2029.
The utility also presented a $26 million project to serve a 300-MW customer in Ashland, Va. A 230-kV substation would be installed on the 230-kV Four Rivers-Hanover line, with about a half-mile of new line required.
Exelon presented a pair of $44 million projects to rebuild two 230-kV lines between the Plymouth Meeting and Whitpain substations in the PECO zone. The conductor on the 5.12-mile lines is about 65 years old, and the tower bolts and paint coatings are 95. The projects are in the engineering phase with one expected to be complete in July 2029 and the other in December 2029.
An additional $54 million project in the PECO zone would replace the 98-year-old Buxmont-Whitpain line. It’s in the engineering phase with a projected in-service date of Dec. 31, 2028.
The utility also presented a need to address limited operational flexibility around transmission outages because of a single 500/230-kV transformer bank being in place at the Limerick substation.
Finally, Exelon presented a $41.5 million project to serve a new service request in the ComEd zone. The customer is expected to bring about 30 MW to the Minooka, Ill., area in June 2028, which could grow to 588 MW by 2036. The project would construct a 345-kV substation, named Wildy Road, with two 345-kV lines to the Kendall County E.C. facility. Five double-circuit lines would feed the customer. The project is in the conceptual phase with a projected in-service date of July 1, 2027.
PJMupdated stakeholders on how it plans to act on a FERC order requiring it to rework how it determines transmission rates and recalculate rates going back to June 2015.
The March 6 order rejected a settlement between several transmission owners and PJM to resolve a complaint filed by Neptune Regional Transmission System and Long Island Power Authority (LIPA), which challenged the de minimis exception (EL15-18, et al.).
The decision instead requires the parties to revise PJM’s tariff to eliminate the exception and recalculate over a decade of rates within 90 days. The exception zeroes out the cost assignment for any zone responsible for less than 1% of the flow modeled on a transmission upgrade. The change is not applicable to costs allocated through the load-ratio share basis.
PJM Associate General Counsel Jessica Lynch said the RTO may request additional time to recalculate costs without the de minimis exception, as the order provides only 90 days to rerun over a decade of billing. The RTO is working to identify the baseline reliability projects affected by the order and what will be needed to recalculate their cost assignment.
Solution-based distribution factors (DFAX) are used to determine the full cost for projects less than $5 million and under 500 kV, while for higher-cost and higher-voltage projects, the calculation is split evenly between the load-ratio share basis and solution-based DFAX. Different methods are used if a project is needed to resolve stability violations.
The order denied a second prong of the Neptune/LIPA complaint, which argued PJM’s practice of netting counterflows, paired with the de minimis exception, distorts how the benefits of a project are accounted for when setting cost allocation. The commission also established a paper hearing to explore whether solution-based DFAX should be used when a project is needed to resolve short circuit violations.
Paul Sotkiewicz, president of E-Cubed Policy Associates, asked what implications the rebilling might have on credit and default risk for PJM members.
PJM General Counsel Chris O’Hara said staff does not have a sense of the scale of the rebilling and that it would be inappropriate to speculate on default risk at this time.
